Towards a comprehensive barcode library for arctic life - Ephemeroptera, Plecoptera, and Trichoptera of Churchill, Manitoba, Canada

Abstract

Background: This study reports progress in assembling a DNA barcode reference library for Ephemeroptera, Plecoptera, and Trichoptera ("EPTs") from a Canadian subarctic site, which is the focus of a comprehensive biodiversity inventory using DNA barcoding. These three groups of aquatic insects exhibit a moderate level of species diversity, making them ideal for testing the feasibility of DNA barcoding for routine biotic surveys. We explore the correlation between the morphological species delineations, DNA barcode-based haplotype clusters delimited by a sequence threshold (2%), and a threshold-free approach to biodiversity quantification--phylogenetic diversity.

Results: A DNA barcode reference library is built for 112 EPT species for the focal region, consisting of 2277 COI sequences. Close correspondence was found between EPT morphospecies and haplotype clusters as designated using a standard threshold value. Similarly, the shapes of taxon accumulation curves based upon haplotype clusters were very similar to those generated using phylogenetic diversity accumulation curves, but were much more computationally efficient.

Conclusion: The results of this study will facilitate other lines of research on northern EPTs and also bode well for rapidly conducting initial biodiversity assessments in unknown EPT faunas.

Figure 1

Circular Neighbour-Joining tree for unique haplotypes of Trichoptera analyzed in this study. A total of 293 haplotypes represent 1500 COI sequences and 68 morphospecies of Trichoptera. Terminal branches of species with intraspecific divergence greater than 2% are highlighted in red. Species groups with subtle diagnostic characters or females/immatures that are difficult to identify morphologically are highlighted in green blocks. Members of these groups can easily be confused or neglected in routine morphological sorting, but are readily detected via barcoding. Red and pink bars in the outer circles represent the maximum intraspecific divergence and minimum distance to nearest neighbour, respectively, of the corresponding species shown in the circular tree. The heights of the two distance bars are proportional to the distance values (see Table S3 in Additional file 3). Two possible cryptic caddisfly species, Limnephilus sansoni and Cheumatopsyche campyla (Ross), are highlighted in yellow blocks.

Figure 2

Circular Neighbour-Joining tree for unique haplotypes of Ephemeroptera analyzed in this study. A total of 123 haplotypes represent 564 COI sequences and 37 morphospecies of Ephemeroptera. Figure symbols and annotations follow those in Fig. 1.

Figure 3

Circular Neighbour-Joining tree for unique haplotypes of Plecoptera analyzed in this study. A total of 44 haplotypes represent 213 COI sequences and 7 morphospecies of Plecoptera. Figure symbols and annotations follow those in Fig. 1.

Figure 4

Barcode haplogroups (2% threshold) compared to morphospecies assignments in species with >2% intraspecific divergence in COI barcodes (in all other cases, the two delineation methods agreed with each other). Clades of focal species were detached from the original haplotype NJ trees presented in Figs. 1, 2 and 3 and reassembled to a pruned tree using the Interactive Tree of Life [29]. The number in parentheses next to the haplotype name indicates the number of sequences sharing the haplotype. The square brackets to the right of the tree represent the haplogroup delineated by a 2% threshold.

Figure 5

Randomized taxon accumulation curves based on molecular and morphological taxon delineations and total phylogenetic diversity. Taxon accumulation curves based on barcode haplogroups and morphospecies were constructed in EstimateS V.8.0 using 50 random replicates and default settings. The phylogenetic diversity (PD) accumulation curve was constructed in R using the packages APE and CAIC (see Methods and Additional file 4). PD values were multiplied by a scaling factor so the end points of the PD and barcode threshold curves would be the same, aiding comparison of the shapes of these curves. As more Churchill EPT individuals are sampled, biodiversity accumulates at a similar rate regardless of how it is quantified, suggesting that all methods applied here could be used in biodiversity assessment. Barcode clusters are more rapidly obtained and efficiently analyzed than morphospecies or total PD.